Micro-assembled devices offer the promise of an entirely new generation of consumer, professional, medical, military, and other products having features, capabilities and cost structures that cannot be provided by products that are formed using conventional macro-assembly and macro-fabrication methods. For example, there is a need, particularly in the field of flat panel displays, smart cards and elsewhere, for microelectronic devices or chips that can be integrated into or assembled as either a system or as an array, in a relatively inexpensive manner. In another example, there is a need for a cost effective method for allowing accurate and cost effective assembly of colored display elements such as electrophoretic beads in specific locations on display panels.
One advantage of such micro-assembled devices is that they can utilize different materials and devices (a process generally termed heterogeneous integration) in ways that create new product possibilities. For example, such heterogeneous integration provides the opportunity for relatively rigid structures such as such as silicon transistors or other electronic devices to be assembled into more complex electronic circuits using a flexible substrate as opposed to the rigid silicon substrates currently used for this purpose. In this example, such heterogeneous integration would provide a less expensive means to assemble silicon based integrated circuit components and/or any other kind of circuit components to form integrated circuits on flexible or rigid supports that are not made from silicon. However, it will be appreciated that in providing such heterogeneous integrated circuits, it is necessary that these processes provide for precise placement of multiple types of independent structures on the substrate. Such heterogeneous integration can also be used for other purposes. For example, heterogeneous integration can be used for purposes such as the assembly of pharmaceutical products, advanced materials, optical structures, switching structures, and biological structures.
Of particular interest in the electronic industry is the potential for micro-assembly to solve existing problems in the assembly of highly desirable but complex structures such as electronic displays. Typical electronic displays use a structure known as a “front plane” as the image forming surface. The “front plane” comprises an arrangement of image forming elements also known as active elements formed from structures such as liquid crystals, electroluminescent materials, organic light emitting diodes (OLEDs), up converting phosphors, down converting phosphors, light emitting diodes, electrophoretic beads, or other materials that can be used to form images. Such active elements typically form images when an electric field or some other stimulus or other field is applied thereto. Such electronic displays also incorporate a structure known as a “back plane” that comprises structures such as electrodes, capacitors, transistors, conductors, and pixel drivers and other circuits and integrating components that are intended to provide appropriate stimulus to the active components to cause the active components to present an image. For example, the active components can react to stimulus by emitting controlled amounts of light or by changing their reflectivity or transmissivity to form an image on the front plane.
It is well known to use heterogeneous integration methods to place elements on a substrate. Such heterogeneous integration methods can be generally divided into one of two types: deterministic methods and random methods. Deterministic methods use a human or robotic structure to place individual elements into particular locations on the substrate. Such methods are also known as “pick and place” methods. Such “pick and place” methods offer two advantages: complete control and positive indication that components have been appropriately placed in a desired location. Further, such “pick and place” methods also allow the precise assembly of different types of micro-components to form a micro-assembled structure that integrates different types of materials, micro-assembled structures and components.
It will be appreciated that deterministic methods require a high degree of precision by the person or machine executing the deterministic assembly process. Accordingly, such deterministic methods are difficult to apply in a cost effective manner. This is particularly true where the assembly of micro-components is to occur at a high rate of assembly or where large-scale assembly of micro-components is to be performed such as is required in commercial, pharmaceutical, or other applications.
Random placement methods such as fluidic self-assembly have been used to integrate electronic devices such as GaAs LEDs onto silicon substrates. Fluidic self-assembly is a fabrication process whereby a large number of individual shaped micro-assembled structures are integrated into correspondingly shaped recesses on a substrate using a liquid medium for transport. This method of self-assembly relies on gravitational and shear forces to drive the self-assembly of micro-components. Examples of this include U.S. Pat. No. 5,545,291 filed by Smith et al. on Dec. 17, 1993 entitled Method for Fabricating Self-Assembling Micro-Assembled Structures; U.S. Pat. No. 5,783,856 filed by Smith et al. on May 9, 1995 entitled Method for Fabricating Self-Assembling Micro-Assembled Structures; U.S. Pat. No. 5,824,186 filed by Smith et al. on Jun. 7, 1995 entitled Method and Apparatus for Fabricating Self-Assembling Micro-Assembled Structures; and, U.S. Pat. No. 5,904,545 filed by Smith et al. on Jun. 7, 1995 and entitled Apparatus for Fabricating Self-Assembling Micro-Assembled Structures.
FIG. 1a illustrates, generally, the operation of one type of prior art random placement method. In FIG. 1a, a substrate 10 is shown having binding sites in the form of recesses 21 that are shaped to accept correspondingly shaped micro-components 47 suspended in a fluid 29. As is shown in FIG. 1a, fluid 29 contains micro-components 47 and is applied to substrate 10. When this occurs, gravity and/or other forces draw micro-components 47 onto substrate 10 and into recesses 21. This allows for the assembly of micro-components 47 to substrate 10 using a massively parallel process that is more suitable for high volume and/or large scale assembly processes.
Other approaches have been developed for using fluidic self-assembly to build a micro-assembled structure without relying exclusively on gravitational and/or shear forces. Some of these are illustrated in FIGS. 1b-1e. In each of FIGS. 1b-1e, a substrate 10 is shown having binding sites 21-25. Binding sites 22-25 can take many forms, only some of which are shown in FIGS. 1b-1e. 
In FIG. 1b, a fluidic self-assembly method is shown wherein a substrate 10 is provided having binding sites 22 that are adapted with hydrophobic patches that engage with hydrophobic surfaces 48 on micro-components 49 suspended in fluid 29 and thereby locate the micro-components 49 on substrate 10. One example of this type is shown and described in U.S. Pat. No. 6,527,964 filed by Smith et al. on Nov. 2, 1999 entitled “Method and Apparatuses for Improved Flow in Performing Fluidic Self-Assembly.” The '964 patent describes a substrate that is exposed to a surface treatment fluid to create a surface on the substrate that has a selected one of a hydrophilic or a hydrophobic nature. A slurry is dispensed over the substrate. The slurry includes a fluid and a plurality of the micro-components. Two types of micro-components are provided: one that is designed to adhere to a hydrophilic surface associated with a co-designed receptor site and one that is designed adhere to a hydrophobic surface associated with a co-designed receptor site. As the slurry is dispensed over the substrate 10, the selectively hydrophilic surfaces of selected ones of the micro-components adhere to hydrophilic surfaces on substrate 10, while not adhering to hydrophobic surfaces. Micro-components that have a hydrophilic surface engage hydrophilic patches on the substrate. Thus, micro-components are selectively placed in predefined locations on the substrate.
FIG. 1c shows another fluidic self-assembly method. The method illustrated in FIG. 1c uses capillary forces for self-assembly. As is shown in FIG. 1c, binding sites 23 are adapted with drops 32 of a liquid 34. Capillary attraction between liquid 34 and surface 36 on micro-components 51 causes micro-components 51 suspended in fluid 29 to assemble on binding sites 23. However, it will be appreciated that this method requires the precise placement of drops of liquid 34 on substrate 10 and does not necessarily provide the discrimination useful in the assembly of components having multiple types of micro-components. Various versions of this method are described generally in Tien et al. (J. Tien, T. L. Breen, and G. M. Whitesides, “Crystallization of Millimeter-Scale Objects with Use of Capillary Forces,” J. Amer. Chem. Soc., vol. 120, pp. 12 670-12 671, 1998.) and Srinivasan et al. (U. Srinivasan, D. Liepamann, and R. T. Howe, J. Microelectromechanical systems Vol. 10, 2001, pp. 17-24).
In the prior art illustrated in FIG. 1d, a fluidic self-assembly method is shown wherein binding sites 24 include magnetic patches that attract a magnetic surface 53 on micro-component 52 suspended in fluid 29. Such an approach is described in Mukarami et al. (Y. Murakami, K. Idegami, H. Nagai, A. Yamamura, K. Yokoyama, and E. Tamiya, “Random fluidic self-assembly of micro-fabricated metal particles,” in Proc. 1999 Int. Conf Solid-State Sensors and Actuators, Sendai, Japan, Jun. 7-10, 1999, pp. 1108-1111.) which describes in greater detail the use of magnetic forces to assemble microscopic metal disks onto a substrate patterned with arrays of nickel dots. However, high cost is encountered in providing the arrays of disks on the substrate. Further such methods are typically limited to applications wherein the micro-assembled structures being assembled each have magnetic characteristics that permit the use of magnetic forces in this fashion.
Electrostatic attraction has been proposed for use in positioning micro-components during micro-assembly. U.S. Patent Publication No. 2002/0005294 filed by Mayer, Jackson and Nordquist, entitled “Dielectrophoresis and Electro-hydrodynamics Mediated Fluidic Assembly of Silicon Resistors”; and S. W. Lee, et al., Langmuir “Electric-Field-Mediated Assembly of Silicon Islands Coated With Charged Molecules”, Volume 18, Pg. 3383-3386, (2002) describe such methods. FIG. 1e illustrates a general example of this electrostatic approach. As is shown in FIG. 1e, substrate 10 has binding sites 25 that are adapted with electrodes 27 that attract oppositely charged micro-components 55 suspended in fluid 29. However, the use of electrostatically based fluidic micro-assembly can involve high cost associated with providing addressable electrode structures required for long range transport of micro-components by dielectrophoresis.
As noted above, many micro-assembled structures incorporate a variety of different types of micro-components. Thus, heterogeneous integration of more than one type of micro-component using such a massively parallel random placement process, such as fluidic micro-assembly, is highly desirable. What is needed therefore is a method for assembling micro-components into a micro-assembled structure on the massive scale enabled by random placement methods such as conventional fluidic assembly but with the precision and selective assembly capabilities of deterministic methods.
Modifications to at least one of the fluidic self-assembly methods described above have been proposed in an attempt to meet this need. For example, in one approach, conventional fluidic assembly techniques have evolved that use differently shaped micro-components that are adapted to engage differently shaped receptor sites on a substrate. This requires that the substrate has binding sites that are uniquely shaped to correspond to a shape of a particular type of micro-component. However, the constraints of surface etching techniques, micro-component formation techniques, cost, electrical function, and orientation limit the number of shape configurations that are available for use in discrimination, which in turn limits the number of different components that can be placed on the substrate using such a process.
In another approach, Bashir et al. discuss the use of binding between complementary DNA molecules or ligands to discriminate between binding sites. While this approach provides a high degree of differentiation high cost may be encountered in patterning the DNA or ligands on the substrate. (H. McNally, M. Pingle, S. W. Lee, D. Guo, D. Bergstrom, and R. Bashir, “Self-Assembly of Micro and Nano-Scale Particles using Bio-Inspired Events”, Applied Surface Science, vol. 214/1-4 pp 109-119, 2003).
Thus, there is a need for a more cost effective method for the high volume heterogeneous assembly of micro-components.